Active Shape Models in the Poultry Processing Industry
Francis Suson Cagatin · Skemman · 2012
The Active Shape Model (ASM) algorithm is an image segmentation technique that takes advantage of the data derived from the training set. The model is built up from shape information obtained from analyzing points along object boundaries known as landmarks. This forms the basis in searching unknown images for target shapes. The theory and process of building up the ASMs and image search is discussed. Examples are given to reinforce the discussion. Finally, the ASM is used on several poultry processing applications. It is found that the inherent ability of ASMs to adapt to different variations seen during training make it an ideal candidate for applications in the poultry processing industry.